library(tidyverse)
library(gapminder)

\(\color{darkblue}{\textbf{Fundamentals}}\)

---

\(\color{dodgerblue}{\textbf{Role of Data Scientist}}\)

---

\(\color{dodgerblue}{\textbf{Modeling is Not Enough}}\)

\(\color{darkblue}{\textbf{Figure Types}}\)

---

\(\color{dodgerblue}{\textbf{Scatter Plot}}\)

Shows the relationship between two variables

\(\color{skyblue}{\textrm{- Basic Plot}}\)

Required: x, y

ggplot(data, 
       aes(x = gdpPercap, 
           y = lifeExp)) + 
  geom_point()

\(\color{skyblue}{\textrm{- Jittering}}\)

geom_point(position="jitter")

ggplot(data, 
       aes(x = continent, 
           y =lifeExp)) + 
  geom_point()

ggplot(data, 
       aes(x = continent, 
           y = lifeExp)) + 
  geom_point(position = position_jitter(width = 0.1))

\(\color{skyblue}{\textrm{- Trend Lines}}\)

geom_smooth(method = "lm")

ggplot(data, 
       aes(x = log(gdpPercap), 
           y = lifeExp)) + 
  geom_point() + 
  geom_smooth(method = "lm")

ggplot(data, 
       aes(x = log(gdpPercap), 
           y = lifeExp, 
           color = continent)) + 
  geom_point() + 
  geom_smooth(method = "lm", se = FALSE) +
  geom_smooth(method = "lm", se = FALSE, 
              aes(group = 1, 
                  color = "Overall Global"),
              linetype=2) +
  scale_color_manual("Region", 
                     values = c("firebrick3", "orange", "#1B9E77", "royalblue3",
                                "mediumpurple4", "black"))

---

\(\color{dodgerblue}{\textbf{Histogram}}\)

Shows the binned distribution of a continuous variable

\(\color{skyblue}{\textrm{- Basic Plot}}\)

Required: x

ggplot(data, 
       aes(x = lifeExp)) +
  geom_histogram()

\(\color{skyblue}{\textrm{- Custom Bins}}\)

Optional: bins or binwidth

ggplot(data, 
       aes(x = lifeExp)) +
  geom_histogram(bins = 3)

ggplot(data, 
       aes(x = lifeExp)) +
  geom_histogram(binwidth = 10)

\(\color{skyblue}{\textrm{- Frequency}}\)

Optional: y (which makes it frequency instead of count)

ggplot(data, 
       aes(x = lifeExp, 
           y = ..density..)) +
  geom_histogram(binwidth = 10)

---

\(\color{dodgerblue}{\textbf{Bar}}\)

Shows the distribution of a discrete variable

\(\color{skyblue}{\textrm{- Absolute Count Bar}}\)

Required: x
Function: geom_bar()

ggplot(data, 
       aes(x = continent)) + 
  geom_bar()

\(\color{skyblue}{\textrm{- Distribution Bar}}\)

Shows a selected statistic for each category of the variable selected

Required: x, y
Function: geom_col()

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp)) %>% 
  ggplot(aes(x = continent, 
             y = mean_life)) + 
  geom_col()

\(\color{skyblue}{\textrm{- Stacked Bar}}\)

fill =

gapminder %>%
  group_by(year, continent) %>%
  summarize(tot_pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = tot_pop, 
             fill = continent)) +
  geom_col()

\(\color{skyblue}{\textrm{- Proportional Fill}}\)

geom_col(position = "fill")

gapminder %>%
  group_by(year, continent) %>%
  summarize(tot_pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = tot_pop, 
             fill = continent)) +
  geom_col(position = "fill")

\(\color{skyblue}{\textrm{- Separate Groups}}\)

geom_col(position = "dodge")

gapminder %>%
  group_by(year, continent) %>%
  summarize(tot_pop = sum(as.numeric(pop))) %>% 
  filter(year %in% c(1957, 1977, 1997)) %>% 
  ggplot(aes(x = continent, 
             y = tot_pop, 
             fill = factor(year))) +
  geom_col(position = "dodge")

\(\color{skyblue}{\textrm{- Reorder Bars}}\)

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = reorder(continent, mean_life), 
             y = mean_life)) + 
  geom_col()

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = reorder(continent, -mean_life), 
             y = mean_life)) + 
  geom_col()

\(\color{skyblue}{\textrm{- Reorder Stack}}\)

gapminder %>%
  group_by(year, continent) %>%
  summarize(tot_pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = tot_pop, 
             fill = continent)) +
  geom_col(position = position_stack(reverse = TRUE)) +
  guides(fill = guide_legend(reverse = TRUE))

\(\color{skyblue}{\textrm{- Bar Size/Spacing}}\)

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = continent, 
             y = mean_life)) + 
  geom_col(width = 0.5)

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = continent, 
             y = mean_life)) +  
  geom_col(width = 1)

---

\(\color{dodgerblue}{\textbf{Line}}\)

Shows a variable over time

\(\color{skyblue}{\textrm{- Basic Plot}}\)

Required: x, y

gapminder %>%
  group_by(year) %>%
  summarize(pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = pop)) + 
  geom_line()

gapminder %>%
  group_by(year) %>%
  summarize(pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = pop)) + 
  geom_line() +
  geom_point()

\(\color{skyblue}{\textrm{- Area Plot}}\)

geom_area()

gapminder %>%
  group_by(year, continent) %>%
  summarize(pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = pop, 
             fill = continent)) +
  geom_area()

\(\color{skyblue}{\textrm{- Proportion Plot}}\)

geom_area(position="fill")

gapminder %>%
  group_by(year, continent) %>%
  summarize(pop = sum(as.numeric(pop))) %>% 
  ggplot(aes(x = year, 
             y = pop, 
             fill = continent)) +
  geom_area(position = "fill")

---

\(\color{dodgerblue}{\textbf{Box}}\)

ggplot(data, 
       aes(x = continent, 
           y = lifeExp)) +
  geom_boxplot()

---

\(\color{dodgerblue}{\textbf{Cleveland Dot}}\)

gapminder %>%
  filter(continent == "Asia", year == 2007) %>% 
  ggplot(aes(x = lifeExp, 
             y = reorder(country, lifeExp))) +
  geom_point(size = 3) +
  theme_bw() +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    panel.grid.major.y = element_line(colour = "grey60", linetype = "dashed"))

---

\(\color{dodgerblue}{\textbf{Functions}}\)

my_fun <- function(xvar) {
  1 / (1 + exp(-xvar + 10))
}

ggplot(data.frame(x = c(0, 20)), 
       aes(x = x)) +
  stat_function(fun = my_fun, 
                geom = "line")

\(\color{darkblue}{\textbf{Aesthetics}}\)

---

\(\color{dodgerblue}{\textbf{Controls}}\)

Aesthetic	Description	Values
x	x-axis data
y	y-axis data
color	color of dots, outline of other shapes
fill	fill color	Same as color
alpha	transparency
size	diameter of points, thickness of lines
linetype	line dash pattern
labels	text on a plot or axes	see below
shape	shape of dot

\(\color{skyblue}{\textrm{- Setting}}\)

Optional: aplpha, color, fill, shape, size

ggplot(data, aes(x = gdpPercap, 
                 y = lifeExp, 
                 fill = continent, 
                 size = pop)) + 
  geom_point(color = "black", 
             alpha = 0.7, 
             shape = 25) 

\(\color{skyblue}{\textrm{- Color Palettes}}\)

Brewer

scale_color_brewer(palette = "Spectral")
scale_fill_brewer(palette = "Spectral")

ggplot(data, aes(x = gdpPercap, 
                 y = lifeExp, 
                 color = continent)) + 
  geom_point() +
  scale_color_brewer(palette = "Spectral")

Manual

scale_color_manual(values = c("color", "color", "color"))
scale_fill_manual(values = c("color", "color", "color"))

ggplot(data, aes(x = gdpPercap,
                 y = lifeExp, 
                 color = continent)) + 
  geom_point() + 
  scale_color_manual(values = c("deeppink", "deepskyblue4", "seagreen4",
                                "darkorchid1", "coral3"))                             

---

\(\color{dodgerblue}{\textbf{Text Location}}\)

\(\color{darkblue}{\textbf{Labels}}\)

---

\(\color{dodgerblue}{\textbf{Adding}}\)

figure + 
  labs(
    title = "This is my title",
    subtitle = "This is my subtitle",
    x = "This is my x-axis",
    y = "This is my y-axis",
    fill = "This is my fill legend",
    color = "This is my color legend",
    caption = "This is my caption"
    )

---

\(\color{dodgerblue}{\textbf{Formatting}}\)

element_text()

• color =
• size =
• face = "plain"/"italic"/"bold"/"bold.italic"
• family =
• hjust =
• vjust =

figure + 
  theme(
    plot.title = element_text(face = "bold.italic"),
    plot.subtitle = element_text(color = "darkgray"),
    plot.caption = element_text(hjust = 0),
    axis.title.x = element_text(hjust = 1)
  )

---

\(\color{dodgerblue}{\textbf{Data Labels}}\)

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = continent, 
             y = mean_life)) + 
  geom_col() +
  geom_text(aes(label = round(mean_life, 2)), 
            vjust = 1.5, 
            color = "white")

\(\color{darkblue}{\textbf{Axes}}\)

---

The scale_ functions include whether the axis is _x_ or _y_ and whether the variable plotted is _discrete() or _continuous():

• scale_x_discrete()
• scale_y_continuous()

---

\(\color{dodgerblue}{\textbf{Limiting}}\)

Caution, this will drop data outside of the limits (including bars)

gapminder_figure + 
  scale_y_continuous(limit = c(30, 100))

---

\(\color{dodgerblue}{\textbf{Setting Breaks}}\)

Caution, this will still plot points outside of the range of the breaks leaving the rest of the y axis blank

gapminder_figure + 
  scale_y_continuous(breaks = c(40, 60, 80))

---

\(\color{dodgerblue}{\textbf{Setting Labels}}\)

Caution, the numbero f labels must equal the number of breaks

gapminder_figure + 
  scale_y_continuous(breaks = c(40, 60, 80),
                     labels = c("Low", "Medium", "High")) 

scales formats:

• comma
• percent
• scientific

gapminder_figure + 
  scale_x_continuous(labels = scales::comma)

---

\(\color{dodgerblue}{\textbf{Scaling}}\)

gapminder_figure + 
  scale_x_continuous(trans = "log10")

\(\color{darkblue}{\textbf{Error Bars}}\)

---

geom_errorbar()

data %>%
  group_by(continent) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = continent, 
             y = mean_life)) + 
  geom_col(fill ="grey50") + 
  geom_errorbar(aes(ymin = (mean_life - sd_life), 
                    ymax = (mean_life + sd_life), 
                    width = 0.2))

gapminder %>%
  filter(country %in% c("Brazil", "Russia", "India", "China")) %>%
  group_by(year) %>%
  summarize(mean_life = mean(lifeExp), 
            sd_life = sd(lifeExp)) %>%
  ggplot(aes(x = year, 
             y = mean_life)) + 
  geom_point() +
  geom_errorbar(aes(ymin = (mean_life - sd_life), 
                    ymax = (mean_life + sd_life), 
                    width = 0.2))

\(\color{darkblue}{\textbf{Annotation}}\)

---

\(\color{dodgerblue}{\textbf{Text}}\)

gapminder_figure + 
  annotate("text", 
           x = 14000, 
           y = 57, 
           label = "This Country",
           vjust = 0,
           hjust = 0,
           color = "darkgreen",
           size = 3,
           fontface = "bold.italic")

---

\(\color{dodgerblue}{\textbf{Lines}}\)

gapminder_figure + 
  geom_hline(yintercept = 60,
             color = "blue",
             linetype = 2) + 
  geom_vline(xintercept = 25000,
             color = "purple",
             linetype = 1)

---

\(\color{dodgerblue}{\textbf{Segments}}\)

gapminder_figure + 
  annotate("segment",
           x = 10000,
           xend = 15000,
           y = 47,
           yend = 56)

---

\(\color{dodgerblue}{\textbf{Shapes}}\)

gapminder_figure + 
  annotate("rect",
           xmin = 9000,
           xmax = 15000,
           ymin = 47,
           ymax = 58,
           alpha = 0.3)

---

\(\color{dodgerblue}{\textbf{Ressession Example}}\)

recessions <- tibble(
  begin = c("1969-12-01", "1973-11-01", "1980-01-01", 
            "1981-07-01", "1990-07-01", "2001-03-01"
            ),
    end = c("1970-11-01", "1975-03-01", "1980-07-01", 
            "1982-11-01", "1991-03-01", "2001-11-01")
  ) %>%
  mutate(begin = as.Date(begin),
         end = as.Date(end)
         )

ggplot(economics, aes(x = date, y = unemploy/pop)) +
  geom_rect(data = recessions, inherit.aes = FALSE,
            aes(xmin = begin, xmax = end, ymin = -Inf, ymax = Inf),
            fill = "red", alpha = 0.2) +
  geom_line() +
  labs(
    x = "",
    y = "Unemployment Rate",
    title = "Relationship Between Recessions and Unemployment",
    subtitle = "Red boxes represent recessions")

\(\color{darkblue}{\textbf{Faceting}}\)

---

facet_wrap(~)
facet_grid(~)

gapminder_figure +  
  facet_wrap(~ continent)

gapminder_figure + 
  facet_grid(~ continent)

gapminder_figure + 
  facet_grid(continent ~ .)

gapminder_figure_all +  
  facet_grid(continent ~ year)

See section on Patchwork for more

\(\color{darkblue}{\textbf{Patchwork}}\)

---

\(\color{dodgerblue}{\textbf{Installation}}\)

# install.packages("devtools")
# devtools::install_github("thomasp85/patchwork")

library(tidyverse)
library(patchwork)

\(\color{dodgerblue}{\textbf{Layout}}\)

\(\color{skyblue}{\textrm{- Basic}}\)

p1 + p2

\(\color{skyblue}{\textrm{- Setting Columns and Rows}}\)

p1 + p2 + p3 + plot_layout(ncol = 1)

p1 + p2 + p3 + plot_layout(nrow = 1)

\(\color{skyblue}{\textrm{- Advanced Layouts}}\)

(p1 | p2 | p3) /
      p4

p1 + (p2 + p3) + p4 + plot_layout(ncol = 1)

\(\color{skyblue}{\textrm{- Nesting}}\)

p4 + 
  {p1 + 
      {p2 + p3 + plot_layout(ncol = 1)}
  } + 
  plot_layout(ncol = 1)

---

\(\color{dodgerblue}{\textbf{Visual Settings}}\)

\(\color{skyblue}{\textrm{- Setting Sizes}}\)

heights(row1, row2, . . .)
widths(col1, col2, . . .)

p1 + p2 + p3 + p4 + plot_layout(nrow = 2, heights = c(1,2), widths = c(2,1))

\(\color{skyblue}{\textrm{- Adding Blank Space Between}}\)

p1 + plot_spacer() + p2

\(\color{dodgerblue}{\textbf{Annotating}}\)

\(\color{skyblue}{\textrm{- Basic Titles}}\)

p1 + p2 + plot_annotation(title = "A great plot!", tag_levels = "A")

\(\color{skyblue}{\textrm{- Better Subtitles}}\)

p1 + labs(subtitle = "A. This is Plot 1") +
  p2 + labs(subtitle = "B. This is Plot 2") + 
  plot_annotation(title = "A great plot!")

\(\color{skyblue}{\textrm{- Formatting}}\)

p1 + p2 + 
  plot_annotation(
    title = "Here is a Complexly Formatted, Simple Graph",
    caption = str_glue("Sources: ggplot\n\nNote: This is just basic analysis as an example."),
    theme = theme(
      plot.title = element_text(face = "bold", hjust = 1),
      plot.caption = element_text(face = "italic", hjust = 0)))

\(\color{darkblue}{\textbf{More Tips}}\)